Vercel's AI Gateway reported that open-weight models handled 56% of production tokens in August, up from 36% in July and under 10% last December, while capturing only 14% of estimated spending as clos…
Running an LLM in a browser tab without an API key or cloud bill is now possible, but users must verify the model file, browser exposure, and data leakage before trusting it with private documents. We…
DeepSQL, an open-source database agent for Postgres and MySQL, released v1.2.0 with schema-scoped access policies that enforce permissions in the planner path, not just via prompts, to prevent LLMs fr…
A new study finds that watermarked AI text fools most readers in blind tests, with only 53% of participants correctly identifying watermarked LLM output as AI versus 51% for unwatermarked LLM output, …
WebGPU is now enabled by default in every major browser, making in-browser LLM inference a deployable tier, according to a tutorial on dev.to that builds a health-log analyzer running Llama-3-8B entir…
Indie developer Dexter Lung released agent-work-board, an MIT-licensed template and scripts that use a single markdown file in git to coordinate parallel Claude Code sessions, reviving the 1970s black…
A three-part book series titled 'The Long Detour' on pre-GPU and low-compute machine learning has been published, drafted by a multi-agent system over a corpus of ~5,000 primary papers and directed by…
NVIDIA and academic researchers introduced ASPIRE, a self-improving robotics framework that achieves 31% zero-shot success on LIBERO-Pro long tasks. The system uses a coordinator-actor architecture wi…
Researchers introduced Jenga Inverse Predictor (JIP-2), a GPU-accelerated deep learning framework that reconstructs collapsed stone structures by treating anastylosis as an inverse prediction task. Th…
Together AI presented eight papers at ICML 2026 spanning the full AI stack, from frontier agents to GPU kernels. Their research includes DSGym for standardized data-science agent evaluation, ThunderAg…
Gray Swan co-founders Zico Kolter and Matt Fredrikson discussed the state of AI red teaming, warning that prompt injection and agent vulnerabilities create a new class of security risks that tradition…
A new paper co-authored by 37 researchers from Stanford, Carnegie Mellon, and the University of Michigan argues that the traditional paper format is obsolete in the AI era, citing "narrative tax" and …
An AI company considering whether to include a specific quality—such as prosocial drives, whistleblowing against power concentration, or strict honesty—in a model spec must weigh criteria across four …
Truffle AI (YC W25), a Bengaluru/San Francisco-based developer platform, raised $500K in seed funding from Y Combinator in February 2025 to turn AI agents into plug-and-play APIs. The platform allows …
RamAIn, a YC W26 company building AI agents that automate enterprise workflows by interacting with legacy systems and desktop applications, is hiring its first business hire. The startup, founded by S…